Carpenter, JH. Observations on the Biology and Behavior of Amphicutis

Home , Ophiocoma echinata, Ophionotus victoriae, Ophiothrix, Ophiothrix angulata, Ophiurida

PROCEEDINGS

OF THE

FIFTEENTH SYMPOSIUM

ON THE

NATURAL HISTORY OF THE BAHAMAS

Edited by Robert Erdman and Randall Morrison

Conference Organizer Thomas Rothfus

Gerace Research Centre San Salvador Bahamas 2016

Cover photograph - "Pederson Cleaning Shrimp" courtesy of Bob McNulty

Press: A & A Printing

© Copyright 2016 by Gerace Research Centre. All rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electric or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in written form.

ISBN 978-0-935909-16-6

The 15th Symposium on the Natural History of the Bahamas

OBSERVATIONS ON THE BIOLOGY AND BEHAVIOR OF AMPHICUTIS STYGOBITA, A RARE CAVE BRITTLE STAR (ECHINODERMATA: OPHIUROIDEA) FROM BERNIER CAVE, SAN SALVADOR ISLAND, BAHAMAS

Jerry H. Carpenter Department of Biological Sciences Northern Kentucky University Highland Heights, KY 41099

ABSTRACT room has larger openings over the water to provide more detritus. Cave isopods and mangrove Amphicutis stygobita is the world’s only rivulus fish do not appear to be significant preda- known brittle star species that appears to be a cave tors of A. stygobita. endemic. Pigment is absent in arms and disk. The statement that A. stygobita is a cave This species was discovered by John Winter in endemic is supported by a surprising array of 2009 and described by Pomory, Carpenter, and troglomorphisms including: no body pigment, Winter in 2011. The few individuals available for muted alarm response to light, reduced body size, the current study were collected in January 2011, elongated arm segments, reduced aggregation, January 2013, and June 2013. slow movement (1-2 cm/min primarily with tube Goals of the current study are to: (1) com- feet, rather than by swinging arms forward), and pare Bernier Cave to nearby Lighthouse Cave to extremely slow regeneration rate of 0.03 mm/wk. possibly determine why A. stygobita are found Bernier Cave and its rare cave brittle star deserve only in Bernier Cave, (2) observe interactions be- much more study. tween A. stygobita and two potential predators: cave isopods (Bahalana geracei) and mangrove INTRODUCTION rivulus fish (Kryptolebias marmoratus), (3) compare some biological and behavioral traits of The brittle star Amphicutis stygobita was A. stygobita (especially troglomorphisms) to those found by John Winter in 2009 in a cave discov- of other brittle stars and to those of cave dwelling ered by Don Bernier on San Salvador Island, Ba- species in other taxa, and (4) determine culture hamas. John Winter and I collected additional techniques to keep A. stygobita alive and healthy. specimens in January 2011; Pomory, Carpenter, Amphicutis stygobita are difficult to find and Winter (2011) described the brittle star as a and collect because they are: found only in the new genus and new species. Distinctive physical dark, tiny (disk diameter 3-4 mm), and similar in traits include: no pigment in arms or disk, translu- appearance to the detritus on which they are cent layer of skin raised above surface of arms, found. They are challenging to maintain in the small size (disk diameter 3-4 mm), and short arms laboratory because they are susceptible to light, (to 9 mm) consisting of relatively few segments salinity changes, and elevated temperatures, and that are elongated (Fig. 1). they don’t accept food normally eaten by brittle In March 2014, Susan Hottenrott Spark stars. kindly sent me unpublished information regarding Bernier Cave has a lower salinity range a single brittle star specimen collected from (~14-28 ppt) than Lighthouse Cave (~20-36 ppt); Lighthouse Cave by a George Washington Uni- it is rare for an echinoderm to be endemic to such versity class in 1992. She still has the specimen, a low salinity environment. Tidal fluctuations and which she studied extensively, and her detailed tidal flow are less in Bernier Cave, allowing for a drawings and description confirm that it is clearly substrate of detritus and soft mud, and its entrance an A. stygobita. Many subsequent searches by

31 The 15th Symposium on the Natural History of the Bahamas

caves without a surface connection to the ocean are called anchialine (Stock et al., 1986). Typical anchialine caves are stratified with a surface freshwater layer from rain water lying over a saltwater layer from the sea, and separated by a halocline layer of mixed fresh water/salt water. Cave divers often swim through several meters of fresh water and halocline to reach the salt water. The anchialine caves on San Salvador Is- land do not have this stratification. Instead, they have salt water or brackish water all the way to the surface. This allows researchers to collect and observe a unique variety of marine cave animals without SCUBA diving. Examples include several new species described from Lighthouse Cave such as the cirolanid isopod crustacean Bahalana Figure. 1. Amphicutis stygobita with metric ruler geracei (Carpenter, 1981), the asselote isopod and detritus. Neostenetroides stocki (Carpenter and Magniez, 1982), and three species of sponges (van Soest George Washington University classes, as well as and Sass, 1981). Another interesting example by my own classes from Northern Kentucky Uni- from San Salvador Island is the remipede crusta- versity, have never found another brittle star in cean Speleonectes epilimnius (Yager and Carpen- Lighthouse Cave. This may indicate that this spe- ter, 1999), described from Major’s Cave; this is cies has been extirpated from the cave, possibly as the only remipede species known to occur at the a result of disturbance from numerous classes that saltwater surface of a cave (hence the name “epi- visit the cave every year. It is also possible that limnius”), rather than several meters deep in strat- the microhabitats in Lighthouse Cave are simply ified cave waters. not as favorable for A. stygobita as those in Ber- Some intriguing questions arose with the nier Cave. discovery of the world’s first cave brittle star. For The discovery of this species is especially instance, does Bernier Cave have unusual quali- noteworthy because it is the only brittle star that ties that allow A. stygobita to live there? And appears to be a cave endemic (Pomory et al. why have we never found A. stygobita in nearby 2011). There are only rare examples of ocean Lighthouse Cave, which my students, colleagues, dwelling species (e.g., Ophioderma longicauda and I have studied nearly every year since 1978? and Ophionereis sp.) finding their way into caves Thus, one of the goals of the current study was to (Pomory et al., 2011). According to Stöhr and compare Bernier Cave to Lighthouse Cave to pos- O’Hara (2013) there are over 2,100 species and sibly determine why A. stygobita are now found subspecies of ophiuroids (brittle stars and basket only in Bernier Cave. stars). Since they live in a large variety of habi- One of my early observations in Bernier tats and are generally photonegative, it is surpris- Cave was that cave isopods, Bahalana geracei, ing that more brittle stars have not been found in are rare there, while mangrove rivulus fish, Kryp- caves. In fact, it seems likely that more brittle tolebias marmoratus (Poey), are fairly common. stars do live in marine caves, but perhaps cave From years of personal observation of Lighthouse divers have not seen or collected these small ben- Cave organisms, I found that both these species thic animals because divers tend to stay off the prey on a variety of animals. Thus, a second goal bottom to avoid clouding the water. of the current study was to observe interactions Limestone caves that are close to oceans between these two potential predators and often contain salt water or brackish water. Those A. stygobita.

32 The 15th Symposium on the Natural History of the Bahamas

Several traits are often associated with MATERIALS AND METHODS cave dwellers, especially with obligate cave organisms known as troglobites (or stygobites, if Animal Sampling aquatic). These special traits are called troglomorphisms (Christiansen, 1962; Romero, 2009). All A. stygobita specimens were collected Familiar examples include loss of eyes and body from dark areas of Bernier Cave (24° 05’ 37” N, pigment; Romero (2009) lists 20 others. It is 7° 27’ 15” W) in shallow water 10-40 cm deep. tempting to think of troglomorphisms as adapta- The few individuals available for this study were tions for living in caves, but in many cases we collected on three trips: 6 January 2011 (15 spec- don’t know how (or if) the characters are actually imens collected, but 7 preserved for type series advantageous in this habitat (Romero, 2009), and and microscopic study); 9-11 January 2013 (n = documentation of adaptation in extremely troglo- 10); and 17 June 2013 (n = 4). Kryptolebias morphic cave animals is difficult (Culver et al., marmoratus fish were collected from Bernier 1995). Nevertheless, it is interesting to examine Cave for fecal studies and predation experiments; what troglomorphisms are found in this first cave collection dates and quantities were: 9-11 January brittle star and possibly support the statement that 2013 (n = 5) and 17 June 2013 (n = 5). We also this is a cave endemic species. Thus, a third goal collected detritus and water samples. Cave iso- of the current study was to compare some biologi- pods (B. geracei) were collected from Lighthouse cal and behavioral traits of A. stygobita (especial- Cave for predation studies; collection dates and ly possible troglomorphisms) to those of non-cave quantities were: 8 January 2013 (n = 6) and 18 dwelling brittle stars and to those of cave dwelling June 2013 (n = 7). More information about Ber- species in other taxa. In order to accomplish this nier Cave and Lighthouse Cave will be covered in third goal it was necessary to pursue a forth the section comparing these two caves. goal—to determine culture techniques to keep Flashlights were used to find and collect A. stygobita alive and healthy long enough to specimens in dark areas of the caves. In January study some of these traits. 2011 aquarium nets were used to collect A. stygo- Amphicutis stygobita is one of the rarest bita from the substrate, but this stirred up detritus brittle star species in the world (Pomory et al., making it difficult to find other specimens, and 2011). On several collecting trips to Bernier Cave arm spines became entangled in the nets. On later no specimens were seen or collected. Since it has trips, fingers or spatulas were used to scoot spec- been impossible to collect A. stygobita in large imens into 35 mm film canisters or small clear numbers, and this species is challenging to keep jars. If more than one specimen was placed in a alive for extended periods, sample sizes have been container it was more likely for oxygen to be de- inadequate for traditional quantitative experi- pleted and arms to be entangled and damaged, so ments. Instead, I am herein describing my obser- we learned to transport specimens in separate con- vations on what might be thought of as early trials tainers. Kryptolebias marmoratus were collected in long-term research projects. I plan to conduct with aquarium nets and transported in individual additional trials and experiments if and when jars. Bahalana geracei were collected in Light- more specimens become available. Nonetheless, house Cave with aquarium nets and transported in it seems appropriate to report my findings now individual jars to avoid cannibalism. Salinities because of the rareness of the species and the were determined with hydrometers and/or refrac- uniqueness of cave endemism for a brittle star. tometers. Temperatures were measured using a After the “Materials and Methods” sec- Weston metal probe thermometer and/or a Suunto tion, this paper is divided into several parts. Each Core wristwatch thermometer. part describes one or more sets of experiments Soon after specimens were collected they and/or observations, results, and my interpretation were taken to the Gerace Research Centre for of observations. short-term observation. Later some were transported in small jars to Kentucky for long-term ob-

33 The 15th Symposium on the Natural History of the Bahamas servation and experimentation. Some A. stygobita camera’s built-in flash was dimmed by setting it were placed in Petri plates or other shallow con- on half power and using a diffuser to reduce pos- tainers for brief observation with an Olympus dis- sible disturbance or damage to specimens. secting microscope. Most A. stygobita specimens Specimens were kept near Bernier Cave were photographed using a Nikon D80 camera temperatures of 23-25°C (73-77°F). I used either with a 60 mm Micro Nikkor macro lens. a 10-gallon tank with a heating pad below, or a water bath with an aquarium heater. Jars of Ber- Culture Methods nier Cave detritus were also kept in darkness at cave temperatures and salinities to try to keep Since I have successfully kept individual them viable and to avoid temperature or salinity specimens of several other cave species (e.g., iso- shock when detritus was added to specimen jars. pods, sponges, and freshwater planarians) alive for long periods (B. geracei for up to eight years), I used the same general methods for A. stygobita. COMPARING BERNIER CAVE TO However, I learned through trial and error that this LIGHTHOUSE CAVE species is more sensitive to light, salinity changes, and higher temperatures so the following methods Since A. stygobita has been found several were more successfully used for A. stygobita. times in Bernier Cave and only once (in 1992) in Each specimen was kept in a small jar with ~30 the extensively studied Lighthouse Cave only ml of brackish water (20-28 ppt), similar to that of ~300 m away, a comparison of the two caves Bernier Cave. Depth in each jar was kept shallow might help explain why. The geology and hy- (1-2 cm) for a high surface area to volume ratio to drology of Lighthouse Cave have been described keep oxygen levels high. Jars had tight-fitting lids by Mylroie (1980) and Davis and Johnson (1988) (rather than loose lids as with Petri plates) to limit respectively, while many pertinent features of evaporation that increases salinity. Salinities were Bernier Cave were described by Pomory et al., checked weekly with a refractometer. If a salinity (2011). The hydrology of Bernier Cave is cur- increased, I reduced it by adding 1-2 ml of water rently being studied by Devin McGinty (a gradu- at a slightly lower salinity; drops were added ate student at University of New Haven) using slowly and away from specimens to avoid salinity data loggers to determine long-term tidal fluctua- shock. A thin layer of detritus from Bernier Cave tions and salinities (pers. comm.). The following was kept on the bottom of each jar to provide food comparisons are based mostly on my personal ob- and to help stabilize levels of oxygen and salinity. servations of these two caves. An additional 8-15 drops of new Bernier Cave Both caves are in the northeastern part of detritus were added every 4-7 days. Animals ig- San Salvador Island, ~1-1.5 km from the ocean. nored or rejected other foods offered including They are reached by taking different paths from small pieces of Tetra Min© fish flakes, shrimp, Dixon Hill Lighthouse (southerly for Lighthouse boiled egg, and boiled lettuce. Each specimen jar Cave and southwesterly for Bernier Cave). Tidal was labeled so I could keep individual records of fluctuations in the caves occur through complex maintenance, experiments, and general observa- conduit systems; tides are delayed (compared to tions. Animals were kept in darkness except dur- ocean tides) by ~45 minutes in Lighthouse Cave ing observations and maintenance. Observations and ~2 hours in Bernier Cave, presumably be- and maintenance were done at night or in a room cause Bernier Cave is further from the ocean. without windows to avoid even weak sunlight, Tidal fluctuations (changes in water depth be- and jars were shaded from direct overhead lights tween low and high tide) are much less and flow and microscope lights. rates are much weaker in Bernier Cave, which Photographs of some A. stygobita individ- help provide a substrate of detritus and soft mud; uals were shot approximately weekly to record detritus appears to be the main source of food for feeding activities and regeneration progress. The A. stygobita. In Lighthouse Cave the substrate is

34 The 15th Symposium on the Natural History of the Bahamas bare rock where tidal water moves rapidly, and it I believe the differences in these two caves is silty in protected quiet water; also, the water that are most likely to explain why A. stygobita and substrate are frequently disturbed by touring are found in Bernier Cave and not in Lighthouse field classes and vacationers. Cave are: salinity range is lower in Bernier Cave, Both caves are entered through openings and slower flow rates and larger openings over the in the roof by climbing down ladders ~3-4 m to water result in abundant detritus on the substrate dry breakdown platforms below. Both caves have for A. stygobita to eat. additional roof openings that permit entrance of sunlight, rain, leaf litter, and fauna including insects, birds, and bats. However, the ceiling open- POTENTIAL PREDATORS OF ings in Bernier Cave are larger and directly above CAVE BRITTLE STARS or near the water, so more detritus enters Bernier Cave water. The entrance room also has much Cave Isopods, Bahalana geracei more light so that mangrove rivulus fish can more easily see and attack prey such as insects, cope- One other possible explanation for why pods, and cave isopods. This appears to have a A. stygobita are currently found in Bernier Cave major effect on populations of cave isopods, and not in Lighthouse Cave is related to food which are common in Lighthouse Cave and rare chains. In previous studies my marine biology in Bernier Cave (only two have been seen in >60 students and I observed predators from Light- hours of observation by researchers). house Cave, and we found that mangrove rivulus There is a significant difference in salinity fish K. marmoratus readily eat B. geracei isopods, ranges of the two caves. Lighthouse Cave is usu- which eat a variety of other cave animals. Bernier ally near full ocean salinity (35-36 ppt), but some- Cave’s well-lit entrance should make it easy for times it drops to ~20 ppt or lower after heavy rivulus fish to prey on B. geracei, which may ex- rains (my samples on 18 June 2013 varied from plain the rareness of the isopods in this cave. If 20-32 ppt depending on location in the cave). Sa- the larger population of B. geracei in Lighthouse linity readings in Bernier Cave range from ~14-28 Cave preyed on A. stygobita they could have elim- ppt; these wide variations in readings are due inated them from that cave, while the reduced mostly to strength of recent rains, but they also population of this predator in Bernier Cave could vary with depth (my samples on 19 June 2013 allow A. stygobita to survive there. The big ques- ranged from 15 ppt at the surface to 21 ppt at 20 tion is, “Will B. geracei prey on A. stygobita?” cm depth near the substrate where brittle stars On five different occasions I put either a healthy were found). or an injured A. stygobita with a B. geracei for It is surprising that A. stygobita can live in several hours or days, and the isopods never such a low salinity, since echinoderms are gener- showed any interest in the brittle stars. To make ally marine because they lack excretory- sure the isopods did not reject the brittle stars osmoregulatory structures (Brusca and Brusca simply because they were not hungry, I offered 1990). In fact, until the discovery of A. stygobita them pieces of shrimp which they ate immediate- in brackish Bernier Cave, the estuarine brittle star ly. On several other occasions I collected another Ophiophragmus filograneus appeared to be the species of small brittle star (Ophiocomella sexra- only endemic brackish-water echinoderm (Turner diata) from an aquarium and placed them with and Meyer, 1980). Since O. filograneus and B. geracei for several days; again the isopods A. stygobita have the unusual ability to tolerate showed no interest in eating them. Since some brackish environments, they may benefit from de- other marine isopods such as the giant Glyptono- creased competition from other echinoderms and tus antarcticus eat brittle stars (Dearborn, 1967; avoidance of predators that cannot tolerate low Janecki and Rakusa-Suszczewski, 2006), it is sur- salinities throughout their life cycles. prising that B. geracei appear to not like the taste or texture of them.

35 The 15th Symposium on the Natural History of the Bahamas

Mangrove Rivulus, Kryptolebias marmoratus Natural Arm Damage

Since brittle stars are often eaten by fish Regeneration is so common in brittle stars (Emson and Wilkie, 1980), I wanted to see if that individuals observed in nature usually have mangrove rivulus fish from Bernier Cave showed regenerating arms at many different growth stages any interest in A. stygobita. On 10 January 2013 (Chinn, 2006). For most species, sub-lethal preda- (one day after they were collected) I added one tion appears to be the major cause of arm damage, A. stygobita to an 8 cm diameter jar with one rivu- so the frequency of regenerating arms is often lus, which soon started attacking tips of the brittle used as an estimate of predation pressure (Clark et star’s arms. I removed the rivulus from the arena al., 2007; Sköld and Rosenberg, 1996). jar to protect the brittle star and examine it for Based on photographs and recorded obser- damage (one arm was now missing the terminal vations of 23 specimens of A. stygobita soon after segment). The next day I returned the rivulus to collection, it appears that the incidence of natural the arena jar with the brittle star, and this time the arm damage is very low compared to that of other rivulus watched the brittle star for ~12 hours but species. Only 8 of 23 specimens (= 35%) showed did not attack it (Fig. 2). I added a second rivulus signs of arm damage or regeneration, and in each to the arena, and both continued to watch but not case only 1 arm showed damage so only 8 of 115 attack for another 12 hours. The arena was mod- arms (= 7%) were damaged. erately well lit for the fish to see, and they had In comparison, Lawrence and Vasques plenty of opportunity to attack, so I concluded that (1996) summarized numerous sub-lethal predation K. marmoratus may not like the taste or texture of studies on 30 species of brittle stars. Some stud- A. stygobita, and probably they are not major ies only described the cause of damage (usually predators of these brittle stars, especially in dark fish predation). Other studies reported the inci- areas of Bernier Cave. This was supported by mi- dence of regeneration in three different ways: (1) croscopic examinations of fecal samples of ten 10 studies of 8 different species reported 43-99% K. marmoratus, which showed no remains of brit- of individuals in populations were regenerating; tle stars. (2) 13 studies of 11 species reported the percent- age of regenerating arms (25-85%); and (3) 6 studies of 5 species simply reported “% of arm regeneration,” without specifying whether % of individuals or % of arms (range <10% to 30- 85%). Thus, in comparison to these studies of at least 21 other species, the incidence of damage is very low in A. stygobita. This indicates that sublethal predation is low and probably has little influence on population density.

TROGLOMORPHISMS

The term “troglomorphism” was proposed by Christiansen (1962) to reflect the convergent morphologies of obligate cave animals (Culver Figure 2. Mangrove rivulus (K. marmoratus) and Pipan, 2009). Some troglomorphisms such as watching Amphicutis stygobita on side of jar. loss of eyes and body pigment seem to be a response to darkness (Romero, 2009). Others, such as slower metabolisms and increases in chemo- and mechanoreceptors to find food are apparent

36 The 15th Symposium on the Natural History of the Bahamas responses to the lack of primary production and ophiuroids have a wide range of coloration” (Po- the resulting low food availability (Romero, mory, 2007; Pomory et al., 2011). 2009). Culver et al., (1995) listed 16 “major characteristics of cave organism,” while Romero Responses to Light (2009) listed 22 troglomorphisms. Of course, “troglomorphy is not universal among subterrane- Although ophiuroids are eyeless, they do an organisms” (Culver et al., 1995), and some have the ability to detect light and to respond to it. troglomorphic characters are specific for certain Brittle stars show a strong tendency to avoid light taxonomic groups. by being nocturnal, which reduces predation by One of my goals in the current study was diurnal fish (Randall, 1967). In fact, light often to examine possible troglomorphisms of A. stygo- acts as an alarm for brittle stars. Reef dwelling bita, which I assembled into Table 1. I omitted brittle stars (e.g., Ophiocoma echinata and Oph- those listed by Culver et al., (1995) and Romero ionereis reticulata) have a strong and immediate (2009) that are not applicable to ophiuroids, in- negative response when a microscope light is cluding reductions in: visual brain centers and shined on specimens resting in the dark (personal pineal organs (absent in ophiuroids), cuticles (pre- observation). In contrast, many times I have sent in terrestrial arthropods), and scales and shined a microscope light on A. stygobita to ob- swimbladders (present in fish). I also omitted serve them, and they generally show little or no several that may apply to A. stygobita, but sup- response, sometimes for >15 minutes. porting information is not yet available, including: Although A. stygobita have relatively little increased egg volume, longer life span, reduced short-term response to light, the overall effect of circadian rhythms, and reduced aggression. I in- light exposure can be severely damaging to brittle cluded the following troglomorphic traits of stars as demonstrated by Johnsen and Kier (1998). A. stygobita that I observed, even if only in anec- In their experiment, after four days of exposure to dotally small numbers; these are discussed in the sunlight, 12 out of 13 Opioderma brevispinum following sections. were dead, and the 13th was damaged; control animals were protected by a UV-opaque filter and Table 1. Troglomorphisms relevant to A. stygo- suffered almost no damage (Johnsen and Kier, bita. 1998). Some cave animals are especially suscep- Eyes and body pigment reduced or absent tible to light; Romero (2009) reported that Bu- Responses to light chanan (1936) exposed the depigmented planari- Reduction in body size an, Sphalloplana percaeca, to direct sunlight and Elongated appendages found that they did not move to shaded areas Enhanced chemoreceptors when given the chance, but they disintegrated Aggregation and fecundity reduced within 12 hours of exposure. On 30 January Slow metabolism: movement and regeneration 2011, I exposed seven A. stygobita to indirect sunlight for 100 minutes to see if they would move away from the light; the data indicated that there Eyes and Body Pigment Reduced or Absent was virtually no directional movement. However, six of the seven specimens died within three days. These two traits are probably the most often This was not a controlled study for effects of UV recognized troglomorphisms. Although A. stygo- light, but the dramatic results convinced me to bita have no eyes, in this case it should not be protect A. stygobita from light. It seems likely considered as troglomorphic since eyes are absent that the translucent layer of raised skin and lack of in virtually all ophiuroids. However, lack of pig- pigment in A. stygobita make them more suscep- ment in the arms and disk or A. stygobita is signif- tible to UV light. icant, since “other shallow-water Caribbean

37 The 15th Symposium on the Natural History of the Bahamas

Reduction in Body Size ing to Brusca and Brusca (1990) sensory neurons in the epidermis of echinoderms “respond to Many cave animals have a reduction in touch, dissolved chemicals, water currents and body size, perhaps as a response to a diminished light.” Thus, it seems likely that A. stygobita may food supply or quality. But Romero (2009) point- also use its raised skin for enhanced chemorecep- ed out that reduction in body size should also be tion and other functions in finding food or mates. advantageous in an ecosystem with an abundance Podia (tube feet) of ophiuroids could also of crevices that could be used for hiding from be considered a type of “appendage,” and Pomory predators, feeding, or reproduction. Amphicutis et al., (2011) noted that the podia in A. stygobita stygobita is one of the smallest brittle stars are relatively large. According to Ruppert and known, with a disk diameter of only 3-4 mm. Barnes (1994) tube feet are used for gas exchange. Pomory (2007) provided a key and approximate Perhaps the enlarged podia (and possibly the skin) disk size for 85 species of ophiuroids from the aid respiration in this warm cave environment Gulf of Mexico and Caribbean Sea; the majority with decaying detritus and no primary production have disk diameters ~10-45 mm, and only two or turbulent water to add oxygen. species have disks <5 mm (Ophiactis algicola ~2 mm, and Ophiactis savignyi ~4 mm.). Thus, it Aggregation and Fecundity Reduced appears that reduction in body size is a troglomorphic trait of A. stygobita. Some ophiuroid species such as Ophio- thrix fragilis occur in dense populations of 1000- Elongated Appendages and Enhanced 2000 per square meter (Ruppert and Barnes, Chemoreceptors 1994). In contrast, most individuals of A. stygobita were found by themselves, but occasionally Culver (1982) pointed out that “the most we have seen two or three within a square meter. cherished tenet of biospeleology” is that “the frag- This low density is probably a reflection of low ile, delicate morphology of cave animals results food quality and/or quantity. Tests have not yet from selection for increased sensory organs on been performed to determine if A. stygobita tend appendages, which in turn results in lengthened to aggregate in captivity. appendages.” But the arms of A. stygobita are Little is known about reproduction and relatively short at only 2-2.5X disk diameter be- fecundity in A. stygobita. However, since juve- cause fewer (but much longer) arm segments are niles have not yet been observed in Bernier Cave, produced. If predation pressure is reduced in it appears that reproduction may be an infrequent Bernier Cave, that would allow A. stygobita to event. conserve energy by moving slower and by pro- ducing fewer arm segments, fewer joints, and Slow Metabolism: Movement fewer spines which are located at the joints. This physical arrangement would result in “stiffer” arm Many studies have found that subterranean movements, possibly making podial walking more species have lower metabolic rates in response to efficient (see section below on “Slow Metabo- food scarcity (Hüppop, 2000). One indication of lism: Movement”). a slow metabolism in A. stygobita is its pattern of The greatly lengthened arm segments may slow movement. Many times I measured the also provide better surface areas for chemorecep- movement of specimens as they slowly glided tion. A distinctive trait of A. stygobita is its trans- around the perimeter of their 5 cm diameter jars. lucent raised layer of skin above the skeletal This usually took 8-15 minutes, yielding a very plates (see Fig. 4 in section below on “Slow Me- slow rate of ~1-2 cm/min. The method of move- tabolism: Regeneration”). Pomory et al., (2011) ment is also unusual. Pomory et al., (2011) point- speculated that the raised skin might be used to ed out that “most ophiuroids move by bending the absorb dissolved organic matter (DOM). Accord- arms, rather than podial walking as is common in

38 The 15th Symposium on the Natural History of the Bahamas asteroids,” but A. stygobita “mainly uses podial walking for locomotion.” Pomory et al., (2011) speculated that “this may be a function of the short arms and the relatively large podia.” It also seems logical that moving by podia (rather than swinging arms) may require less energy, and rapid escapes from predators are apparently needed less frequently in this cave habitat.

Slow Metabolism: Regeneration Rates

Most of my A. stygobita specimens did not lose any arms, or if they lost arms they did not survive long enough to have significant regeneration. However, one specimen (R #1) collected in January 2013 that lost parts of four arms had steady regeneration for 21 weeks; a second specimen (R #2) collected in June 2013 lost parts of Figure 3. Amphicutis stygobita (specimen R #2) all five arms, and it regenerated for 24 weeks. after 5 months of regeneration. Symbols indicate Both specimens probably lost arms as a result of beginning of regenerated arm sections. reduced water quality from increased temperature and decreased oxygen during their long trips home. The two specimens were similar in size (disk diameter 4 mm), both started with arms ~5.5 mm long and with 14 segments/arm, and they had similar patterns and rates of regeneration. They fed regularly on detritus and appeared to be healthy. The number of regenerating segments on each arm was determined from microscopic ex- amination and photographs taken at intervals of 1- 3 weeks; the number of regenerated segments ranged from only 0-3 for R #1 in 21 weeks, and only 1-4 for R #2 in 24 weeks (Figs. 3-4). At first regeneration rates were recorded simply as number of segments added, but that did not correlate well to published regeneration rates for other species, which were typically reported as mm/wk. So, I measured the small growth in each regenerating arm of the two specimens based on photo- Figure 4. Amphicutis stygobita 3rd arm enlarged graphs. Results varied from no growth for two from Figure 3, showing skin and regenerated arm arms to a maximum of 1 mm in 24 weeks for two section (^). arms (= 0.04 mm/wk). The average regeneration rate for all regenerating arms was 0.03 mm/wk. Clark et al., (2007) presented a table of regeneration rates (RR = mm/wk) at various temperatures for 13 species of brittle stars as reported in 10 different studies. The fastest regeneration

39 The 15th Symposium on the Natural History of the Bahamas rate was 6.25 mm/wk at 28°C for Microphiopholis at 25°C should be at least four times faster than gracillima (now known as Amphipholis gracilli- that for O. victoriae, instead of slower. Further- ma), while the slowest regeneration rates were more, the slow regeneration rate of A. stygobita is 0.25 mm/wk at 14°C for Amphipholis squamata contrary to the “general phenomenon that smaller and 0.22-0.68 mm/wk at 0.75°C for the Antarctic species of animals typically have higher per unit brittle star Ophionotus victoriae. So, the regener- mass metabolic rates than larger species” (Pomory ation rate of A. stygobita of 0.03 mm/wk is much and Lawrence, 1999). slower (by a factor of 10) than the slowest rates of the other species presented by Clark et al., (2007). Table 2. Regeneration rates (RR) and relative Since arms of A. stygobita are relatively regeneration rates (RRR) for ten brittle star species short, 2-2.5X disk diameter (Pomory et al., 2011), measured at different temperatures, adapted from theoretically the time required to regenerate an Clark et al. (2007). entire arm should also be fairly short, but at the RRR slow regeneration rate of 1 mm in 5 months it Disk RR (RR/ Temp. Species size Source would take 27.5 months (= 2.3 years) to re-grow a (mm/wk) disk (°C) (mm) 5.5 mm arm, and 45 months (= 3.75 years) to size) Amphicutis This study grow a full-length 9 mm arm. According to Clark 0.03 4 <0.01 25 stygobita et al. (2007), “an Antarctic brittle star with an 80 Amphipholis D’Andrea et 6.25 8 0.78 28 mm arm could take over 3 yr to re-grow a com- gracillima al. (1996) Amphipholis Dupont et al. 0.25 5 0.05 14 plete arm.” These re-growth times are extremely squamata (2001) Amphiura Thorndyke et long compared to fast-growing species such as 0.8-1.5 10 0.11 7-12 filiformis al. (2003) Amphipholis gracillima, which has a regeneration Hemipholis McAlister rate at 28°C that would “indicate complete re- elongata 2.1 10 0.21 20 and Stancyk (2003) placement of lost tissue in 100 to 120d” (Stancyk Macrophi- Chinn (2006) et al., 1994). opholis longi- 6 20 0.30 26-29 peda One deficiency in the Clark table is that it Ophiarthrum Chinn (2006) 3 25 0.12 26-29 does not address the concept that larger species elegans Ophiocoma Chinn (2006) would naturally add more mm/wk than smaller 3.6-4.2 20 0.20 26-29 scolopendrina species. To better compare the regeneration rate Ophionotus Clark et al. 0.22-0.68 20 0.02 0.75 in A. stygobita to that of other species, I collected victoriae (2007) Ophiophrag- Clements et disk diameters from various sources (e.g., Po- mus filogra- 2.94 10 0.29 23 al. (1994) mory, 2007). Then I modified the Clark table into neus my Table 2, which includes ten species, regeneration rates (RR = mm/wk), disk diameters (mm), Several factors may influence the excep- relative regeneration rates (RRR = RR/disk mm), tionally slow growth rate in A. stygobita. First, temperatures, and sources of information. Rela- cave animals often have inherently reduced meta- tive regeneration rates (RRR), which account for bolic rates, possibly as an adaptation to an energy disk size, still show that A. stygobita has the slow- poor environment (Poulson, 1963). A second est regeneration rate (<0.01 mm/wk/mm of disk), factor—one associated with reduced metabolic compared to the next slowest for Ophionotus vic- rate—is that the detritus which A. stygobita eats toriae (0.01-0.03 mm/wk/mm of disk). This dif- would actually seem to be a relatively poor energy ference is even more noteworthy because of the source compared to food given to regenerating dramatically different temperatures (25°C for A. brittle stars in some other studies (e.g., Chinn, stygobita compared to 0.75°C for O. victoriae). 2006 used turkey meat, Clark et al., 2007 used According to the Q10 phenomenon, biological white fish, and Pomory and Lawrence, 1999 used processes in ectotherms roughly double with each TetraMin© fish food). 10°C rise in temperature (Clark et al., 2007). By A third factor to consider is salinity. Spec- this measure the regeneration rate for A. stygobita imens of A. stygobita were maintained at Bernier

40 The 15th Symposium on the Natural History of the Bahamas

Cave salinities of 20-28 ppt. Donachy and and Lawrence, 1999, and the current study) in- Watanabe (1986) found that length of regenerated cluded lag phase days in calculations of regenera- arms and number of ossicles formed in Ophiothrix tion rates. If Clark et al., (2007) had included lag angulata were significantly less at 23 ppt than at phase days, the 0.44 mm/wk regeneration rate for 28-38 ppt; they correlated this with reduced calci- O. victoriae would be reduced by 5/12 (= 43%) um concentrations at lower salinities. resulting in a rate of ~0.25 mm/wk; RRR (relative Fourth, Clark et al., (2007) noted that in regeneration rate) would then be 0.25/20 mm disk their study of the Antarctic brittle star, “the longer size = 0.013 and nearly as low as the RRR calcu- length of arm lost, the faster the regeneration lated for A. stygobita at <0.01. Of course, another rate.” In the Clark study, “one arm was amputat- major factor in my calculations of RRR is disk ed from each individual approximately 10 seg- size itself, and for this I used approximations from ments from the disc;” this is relatively close to the the literature rather than actual disk diameters be- disk in this long-armed species. In some other cause most studies (including Clark et al., 2007) studies (e.g., Zeleny, 1903; Pomory and Law- did not state disk sizes of their regenerating spec- rence, 1999) amputations were made at or near the imens. disk. By comparison, the two A. stygobita speci- To summarize this section on regenera- mens lost only ~28-64% of their total arm length, tion, the regeneration rate of A. stygobita is one of and regeneration rates appeared to slow as arms the slowest ever observed. The most likely con- grew closer to full length. Thus, a short section of tributing factors include: (1) inherently slow me- a damaged arm might be expected to grow slower tabolism related to cave endemism, (2) feeding on than an almost completely removed arm. detritus as a poor quality food, (3) reduced salinity A fifth factor is that several arms were re- and calcium availability, and (4) relatively short generating simultaneously. As Chinn (2006) sections of arms regenerating. pointed out, “it is logical that the more arms a brittle star is regenerating, the slower the growth CONCLUSIONS will be, due to increased energy expenditure.” However, Zeleny (1903) in his study of Ophio- Bernier Cave’s unusual features of brack- glypha lacertosa, found that “the greater the num- ish water, low tidal fluctuations, well-lit entrances ber of removed arms (excepting the case where all open to the surface, and a substrate with abundant are removed) the greater is the rate of regenera- detritus provide a favorable environment for cave tion of each arm.” Thus, I suspect that the simul- brittle stars and other organisms. Cave isopods taneous regeneration of arms in A. stygobita did and mangrove rivulus do not appear to be signifi- not contribute to the slow rate. cant predators of A. stygobita. Special culture A sixth factor is the way lag phases were techniques are required for cave brittle stars be- used in calculating regeneration rates. Clark et cause of restricted food requirements and suscep- al., (2007) noted that O. victoriae had an unusual- tibility to light, salinity changes, and elevated ly long “lag phase of up to 5 mo before reproduc- temperatures. The statement that A. stygobita is a ible amounts of new tissue are produced,” so they cave endemic is supported by its surprising array only plotted regenerated arm lengths for months 5 of troglomorphic traits including: no body pig- to 12. The next longest lag phase found in the ment, muted alarm response to light, reduced literature is “up to one month” for Astrobrachion body size, elongated arm segments, raised skin constrictum at 10.5° to 14.5°C (Stewart, 1996). possibly for enhanced chemoreception and possi- The lag phase for A. stygobita specimen R #2 was bly for DOM absorption, reduced aggregation, at least 26 days; the lag phase of specimen R #1 slow movement, and extremely slow regeneration. appeared to be less than 12 days. According to Bernier Cave and its rare inhabitant, Amphicutis Clark et al., (2007) the lag phase in some species stygobita, deserve much more study by research- “can be minimal, often lasting 2-4 d.” Thus, most ers. Hopefully, this special cave will be protected regeneration studies (e.g., Chinn, 2006, Pomory

41 The 15th Symposium on the Natural History of the Bahamas from class tours and projects that could be detri- Chinn, S. 2006. Habitat distribution and compar- mental to its unique cave brittle stars. ison of brittle star (Echinodermata: Ophiu- roidea) arm regeneration on Moorea, ACKNOWLEDGEMENTS French Polynesia. Water Resources Cen- ter Archives. Biology and Geomorphology I would like to thank Dr. Donald T. of Tropical Islands (ESPM 107/IB 158), Gerace, Chief Executive Officer, and Tom Roth- University of California, Berkeley, CA. fus, Executive Director of the Gerace Research Available at: Centre, San Salvador, Bahamas for sponsoring the http://repositories.cdlib.org/wrca/moorea/ symposium and helping with research logistics; chinn. Rob Erdman for coordinating symposium activities; and Erin Rothfus for facilitating collecting Christiansen. K.A. 1962. Proposition pour la permit applications. Special thanks to John Win- classifycation des animaux cavernicoles. ter for discovering and collecting the cave brittle Spelunca Memoires 2:76-8. stars; to Li Newton for her many hours of cutting trails and assistance in collecting specimens; to Clark, M.S., S. Dupont, H. Rossetti, G. Burns, Chris Pomory for his excellent work on our origi- M.C. Thorndyke, and L.S. Peck. 2007. De- nal 2011 paper and for helpful comments on this layed arm regeneration in the Antarctic paper; and to Susan Hottenrott Spark for provid- brittle star Ophionotus victoriae. Aquatic ing information regarding the single specimen Biology 1:45-53. collected from Lighthouse Cave in 1992. Clements, L.A.J., S.S. Bell, and J.P. Kurdziel. REFERENCES 1994. Abundance and arm loss of the in- faunal brittlestar Ophiophragmus filogra- Brusca, R.C. and G.J. Brusca. 1990. Inverte- neus (Echinodermata: Ophiuroidea), with brates. Sinauer. Sunderland, MA. 922 pp. an experimental determination of regeneration rates in natural and planted seagrass Buchanan, J. 1936. Notes on an American cave beds. Marine Biology 121:97-104. flatworm, Sphalloplana percoeca (Pack- ard). Ecology 17:194-211. Culver, D.C. 1982. Cave Life: Evolution and Ecology. Harvard University Press. Cam- Carpenter, J.H. 1981. Bahalana geracei, n. gen., bridge, MA. 189 pp. n. sp., a troglobitic marine cirolanid isopod from Lighthouse Cave, San Salvador Is- Culver, D.C., T.C. Kane and D.W. Fong. 1995. land, Bahamas. Bijdragen tot de Dierkun- Adaptation and Natural Selection in de 51:259-267. Caves. The Evolution of Gammarus minus. Harvard University Press. Cambridge, Carpenter, J.H. and G.J. Magneiz. 1982. Deux MA. 223 pp. asellotes stygobies des indes occidentales: Neostenetroides stocki n. gen., n. sp., et Culver, D.C. and T. Pipan. 2009. The Biology of Stenetrium sp. Bijdragen tot de Dierkunde Caves and other Subterranean Habitats. 52:200-206. Oxford University Press, Oxford. U.K. 256 pp.

42 The 15th Symposium on the Natural History of the Bahamas

D’Andrea, A.F., S. Stancyk, and G.T. Chandler. Johnsen, S. and W.M. Kier. 1998. Damage due to 1966. Sublethal effects of cadmium on solar ultraviolet radiation in the brittlestar arm regeneration in the burrowing brit- Ophioderma brevispinum (Echinoderma- tlestar, Microphiopholis gracillima. Eco- ta: Ophiuroidea). Journal of the Marine toxicology 5:115-133. Biological Association of the United Kingdom 78:681-684. Davis, R.L. and C.R. Johnson. 1988. Karst hydrology of San Salvador. Pp. 118-135 in Lawrence, J.M., and J. Vasquez. 1996. The ef- Mylroie, J. ed. Proceedings of the 4th Sym- fect of sublethal predation on the biology posium on the Geology of the Bahamas. of echinoderms. Oceanological Acta Bahamian Field Station, San Salvador, 19:431-440. Bahamas. McAlister, J.S., and S.E. Stancyk. 2003. Effects Dearborn, J.H. 1967. Food and reproduction of of variable water motion on regeneration Glyptonotus antarcticus (Crustacea: Iso- of Hemipholis elongate (Echinodermata: poda) at McMurdo Sound, Antactica. Ophiuroidea). Invertebrate Biology Transactions of the Royal Society of New 122:166-176. Zealand. Zoology 8(15):163-168. Mylroie, J.E. 1980. Caves and Karst of San Sal- Donachy, J.E., and N. Watanabe. 1986. Effects vador: Field Guide to San Salvador Is- of salinity and calcium concentration on land, Bahamas, CCFL, Ft. Lauderdale, arm regeneration by Ophiothrix angulata Florida, pp. 67-91 (revised, 1983 for the (Echinodermata: Ophiuroidea). Marine 3rd edition). Biology 91:253-257. Pomory, C.M. 2007. Key to the common shallow Dupont, S., D. Vangeluwe, and J. Mallefet. 2001. water brittle stars (Echinodermata: Arm regeneration in the luminous ophiu- Ophiuroidea) of the Gulf of Mexico and roid Amphipholis squamata: a functional Caribbean Sea. Caribbean Journal of Sci- approach. Pp. 255-259 in Barker, M., ed., ence Special Publication, 10, 42 pp. Echinoderm Research 2000. Swets and Available from: Zeitlinger, Lisse. NL. http://caribjsci.org/epub10.pdf.

Emson, R.H., and I.C. Wilkie. 1980. Fission and Pomory, C.M. and J.M. Lawrence. 1999. Effect of autotomy in echinoderms. Oceanography arm regeneration on energy storage and and Marine Biology - An Annual Review gonad production in Ophiocoma echinata 18:155-250. (Echinodermata: Ophiuroidea). Marine Biology 135:57-63. Hüppop, K. 2000. How do cave animals cope with the food scarcity in caves? Pp. 159- Pomory, C.M., J.H. Carpenter, and J.H. Winter. 188 in H. Wilkens, D.C. Culver, and W.F. 2011. Amphicutis stygobita, a new genus Humphreys, eds. Subterranean Ecosys- and new species of brittle star (Echino- tems, Elsevier Press, Amsterdam. NL. dermata: Ophiuroidea: Ophiurida: Am- philepididae) found in Bernier Cave, an Janecki, T. and S. Rakusa-Suszczewski. 2006. Bi- anchialine cave on San Salvador Island, ology and metabolism of Glyptonotus ant- Bahamas. Zootaxa 3133:50-68. arcticus (Eights) (Crustacea: Isopoda) from Admiralty Bay, King George Island, Antarctica. Polar Bioscience 19:29-42.

43 The 15th Symposium on the Natural History of the Bahamas

Poulson, T.L. 1963. Cave adaptation in ambly- Thorndyke, M.C., M. Patruno, Y. Dewael, S. opsid fishes. American Midland Natural- Dupont, and J. Mallefet. 2003. Regenera- ist 70:257-290. tion in the ophiuroid Amphiura filfiformis: cell biology, physiology and biolumines- Randall, J.E. 1967. Food habits of reef fishes of cence. Pp. 193-199 in Feral, J.P., B. Da- the West Indies. Studies in Tropical vid, eds., Echinoderm Research 2001. Oceanography 5:665-847. Swets & Zeitlinger, Lisse. NL.

Romero, A. 2009. Cave Biology, Life in Darkness. Turner, R.L. and C.E. Meyer. 1980. Salinity toler- Cambridge University Press. Cambridge, ance of the brackish-water echinoderm UK. 291pp. Ophiophragmus filograneus (Ophiu- roidea). Marine Ecology Progress Series Ruppert, E.E. and R.D. Barnes. 1994. Invertebrate Vol. 2:249-256. Zoology. Saunders College Publishing. Philadelphia, PA. 1102 pp. van Soest, R.W.M., and D.B. Sass. 1981. Marine sponges from an island cave on San Sal- Sköld, M., and R. Rosenberg. 1996. Arm regen- vador Island, Bahamas. Bijdragen tot de eration frequency in eight species of Dierkunde 51(2):332-344. Ophiuroidea (Echinodermata) from Euro- pean sea areas. Journal of Sea Research Yager, J. & J.H. Carpenter. 1999. Speleonectes 35:353-361. epilimnius, new species, (Remipedia: Spe- leonectidae) from surface water of an an- Stancyk, S.E., H.M. Golde, P.A. Pape-Lindstrom, chialine cave on San Salvador Island, Ba- and W.E. Dobson. 1994. Born to lose. I. hamas. Crustaceana 72:965-977. Measures of tissue loss and regeneration by the brittlestar Microphiopholis gracil- Zeleny, C. 1903. A study of the rate of regenera- lima (Echinodermata: Ophiuroidea). Ma- tion of the arms in the brittle-star Ophio- rine Biology 118:451-462. glypha lacertosa. Biological Bulletin 6:12-17. Stewart, B. 1996. Sub-lethal predation and rate of regeneration in the euryalinid snake star Astrobrachion constrictum (Echinoderma- ta: Ophiuroidea) in a New Zealand fiord. Journal of Experimental Marine Biology and Ecology 199:269-283.

Stock, J.H., T.M. Iliffe, and D. Williams. 1986. The concept “anchialine” reconsidered. Stygologia 2:90-92.

Stöhr, S. and T. O’Hara, eds. 2013. World Ophiuroidea database. Accessed at http://marinespecies.org/ophiuroidea on 2013-11-12.